Add extra delay to reboot playbook #2128
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request adds a 30-second delay after rebooting a host to prevent connection issues with subsequent tasks. The change is reasonable and directly addresses the CI failures mentioned. I've suggested making the delay configurable by using a variable instead of a hardcoded value, which would improve maintainability.
We often see CI failures where the reboot playbook is successful, but the growroot playbook invoked immediately after it fails with a `Connection timed out` error. Add a small 5 seconds extra delay after the reboot to ensure hosts have finished booting successfully. The delay can be customised with the `post_reboot_delay_s` variable.
priteau
changed the title
|
/gemini review |
There was a problem hiding this comment.
Code Review
This pull request aims to improve the reliability of the reboot playbook by adding a post-reboot delay. This is a sensible addition to address the CI failures described. However, the implementation introduces a critical issue by incorrectly applying an | int filter to the reboot_timeout parameter. This would cause the timeout to be effectively zero, likely leading to more failures. I have left a specific comment with a suggested fix for this issue. The addition of the post_reboot_delay and the corresponding release note are otherwise correct.
2 participants
We often see CI failures where the reboot playbook is successful, but the growroot playbook invoked immediately after it fails with a
Connection timed outerror.Add an extra delay after the reboot to ensure host has finished booting successfully.